SHIN HATI FLUX
Well, this was bound to happen, according to the way times are nowadays, I experimented with FLUX training and it is awesome. Updated model for your enjoyment!
TRIGGER WORD: shinflux
________________________________________________________________________________________________
SHIN HATI XL | LoRA |
XL version of my Shin Hati model (StarWars Ashoka Series), my notebook smokes every time I generate an image, however I had to test and generate a model in XL to see the differences and they are noticeable...
_______________________________________________________________________________________________
SHIN HATI SD1.5 | LoRA |
Well what can I say, this is my second attempt exploring the world of Loras' training, it's been a lot of trial and error, but I think I finally got a relatively good model, I leave it here for the fans of this new character from StarWars Ahsoka to enjoy, I'll be watching to see what you can do with it.
Greetings!
Features:
- Trained in SD1.5.
- Keyword "shtyocta" (without quotes, don't screw it up), though preferably when stating it I prefer to use the lora command with intensity 1, example:
<lora:shtyocta-05:1>
- Play with words for wearing or not a cape or driving a spaceship, etc.
- Tested with realisticVisionV51_v51VAE, epicrealism_naturalSinRC1VAE, cyberrealistic_v33 models, if you can use other models it would be interesting to see the results.
Description
UPDATE TO VERSION 2 !!!
An update of the model, more faithful to the original character, more moldable, more realistic.
Details:
- Generated with Stable-diffusion-v1-5/v1-5-pruned.
- Good results with Checkponit Juggernaut_Aftermarth.
- Heavier but more detailed.